Addresses PR #365 review:
- Type `resolve_sandbox_root`'s resolver param as a `ResolverLike` Protocol
(structural, `resolve_bottle_id` only) — fixes the pyright errors from
passing a duck-typed fake resolver in tests; mirrors
egress_addon_core.PolicyResolverLike.
- Rename `_project_root`/`project_root`/`resolve_repo_root`/
`DEFAULT_PROJECT_ROOT` -> `_sandbox_root`/`sandbox_root`/
`resolve_sandbox_root`/`DEFAULT_REPO_ROOT`: "sandbox" names the per-tenant
scope; "project" was too amorphous. `GIT_PROJECT_ROOT` keeps git's own
env-var name.
- Note the single-tenant path is transitional (stripped once every backend
runs the consolidated gateway).
- policy_resolver: document that the optional identity token is
transitional — the consolidated end state requires it (flips at the
/resolve boundary once token *delivery* lands, a PRD 0070 open question).
- Split the long line in main() (was 105 cols).
pyright 0 errors; pylint 9.95/10; unit suite green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
First slice of git-gate consolidation: make the smart-HTTP git backend
serve every bottle from one process, selecting each request's repo root by
the calling bottle's source IP — the same attribution invariant + resolver
the multi-tenant egress addon uses. `git daemon` can't source-IP-route per
connection, so the consolidated gateway serves git-gate over this HTTP
backend (the transport firecracker/macOS already use); wiring the docker
path onto it lands with the launch-integration slice.
- policy_resolver: factor out `_post_resolve`; add `resolve_bottle_id`
(source IP -> bottle id, same fail-closed 403->None contract as
`resolve`) — the git-gate has no policy blob to parse, the bottle *is*
the namespace.
- git_http_backend: `resolve_repo_root(resolver, base, source_ip, token)`
— single-tenant passthrough when no resolver; else `<base>/<bottle_id>`,
fail-closed on unattributed / resolver error / namespace escape.
`BOT_BOTTLE_ORCHESTRATOR_URL` (same env as egress) flips the shared
gateway multi-tenant; the identity header is read for attribution and
never forwarded to the CGI. Per-repo creds + hooks scope by repo dir, so
isolating the root per bottle isolates its creds too.
Single-tenant path unchanged (existing real-git-push tests green). New unit
coverage for the resolver + repo-root selection matrix. Full unit suite
green (1690 tests; the 13 test_sidecar_init /bin/sleep errors are the
pre-existing NixOS-local noise).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Retire "sidecar" for the consolidated per-host path (PRD 0070 naming
decision): the orchestrator is the umbrella/control plane, and the
egress/git/supervise data-plane unit it runs is the "gateway".
- git mv sidecar.py -> gateway.py and the two integration + one unit test
files; DockerSidecar->DockerGateway, Sidecar->Gateway,
SidecarError->GatewayError, SIDECAR_*->GATEWAY_*, ensure_sidecar->
ensure_gateway, sidecar_status->gateway_status, container name
bot-bottle-orch-sidecar->bot-bottle-orch-gateway.
- Prose rename across broker/registry/egress/policy_resolver + PRD 0070.
- Preserved: the image name bot-bottle-sidecars, the
BOT_BOTTLE_SIDECAR_IMAGE env var, Dockerfile.sidecars, and PRD 0069's
own stage-name cross-references (that doc still uses "sidecar").
No behavior change. Full unit suite green (1679 tests; the 13
test_sidecar_init /bin/sleep errors are pre-existing NixOS-local noise).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The egress addon now selects each request's Config by the calling bottle's
source IP, so one shared sidecar serves every bottle. Opt-in and fail-closed;
single-tenant behaviour is unchanged.
Orchestrator side (source-IP-primary attribution, per the PRD invariant):
* registry: `by_source_ip` (the single active bottle at a source IP —
network-layer attribution); `attribute` now composes it + the token.
* service: `resolve(source_ip, token="")` — with a token, strict
attribution; without, source IP alone.
* control_plane: `POST /resolve`'s identity_token is now OPTIONAL (absent
→ source-IP-only); split cleanly from the token-required `/attribute`.
* policy_resolver: `resolve` token now optional.
Egress side:
* egress_addon_core: `resolve_client_config(resolver, client_ip, token)` —
fetches + parses the client's Config, **fail-closed**: unattributed, a
resolver error, or an unparseable policy all yield deny-all (no routes).
Host-testable; `PolicyResolverLike` Protocol keeps it import-free.
* egress_addon: consolidated mode when `BOT_BOTTLE_ORCHESTRATOR_URL` is
set → `_active_config(flow)` resolves per client IP (reads + strips the
`x-bot-bottle-identity` header); `request()` uses it. Unset → the static
routes file, exactly as before. `PolicyResolver` added to the bundle.
Security note: source-IP-only resolution is safe where the IP is unspoofable
(Firecracker /31 + nft) AND the control plane is reachable only by the
trusted sidecar; the identity token, when the agent injects it, strengthens
it on weaker backends.
Scope note: the egress data plane is now multi-tenant. Remaining to be fully
live: the network topology routing every bottle's proxy to the one shared
sidecar, git-gate multitenancy, and agent-side identity-token injection.
Tests: registry by_source_ip; orchestrator resolve (with/without token);
control-plane /resolve token-optional; resolver token-optional;
resolve_client_config fail-closed matrix. All 182 egress tests still pass
(single-tenant unchanged). Full suite green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Review: the resolver is called rarely enough that a round-trip doesn't
matter, and correctness beats speed — always fetching means a revocation,
policy change, or teardown the orchestrator knows about is honored
immediately instead of lingering for a cache TTL. Remove the TTL cache
(and `invalidate`); `resolve` now hits the orchestrator every call. Noted
in the docstring that any future caching should use orchestrator-driven
invalidation, not a blind TTL.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The data-plane bridge that lets the consolidated sidecar apply each
bottle's policy per request. `PolicyResolver` resolves a client's policy
from the orchestrator's `POST /resolve` keyed on (source_ip, identity
token) and caches it briefly (short TTL) so it isn't a round-trip per
request; `invalidate()` drops an entry on teardown / live reload.
Fail-closed: an unattributed client (orchestrator answers 403) resolves to
None so the caller denies; unreachable / unexpected status raises so the
caller can fail closed too rather than serve stale/empty policy. Stdlib
only and free of bot-bottle imports, so it can be COPYed flat into the
sidecar bundle.
Scope note: this is the sidecar-side *client*. Wiring it into the live
egress mitmproxy addon (select `Config` per client IP in the request path)
and git-gate, plus routing all bottles' egress to the one shared sidecar,
are the remaining data-plane pieces — a heavier change to the sidecar
bundle's adversarial-input code, taken next.
Tests: resolve returns/caches/expires/invalidates; 403 -> None (fail
closed); other HTTP status + unreachable raise; missing policy -> empty;
posts source_ip + identity_token. Full suite green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The orchestrator side of the multi-tenant consolidated sidecar: hold each
bottle's sidecar policy and serve it by verified source IP, with live
reload. One shared sidecar can now get per-bottle config keyed on who's
calling.
* registry.py — a `policy` column (migration v3, opaque JSON the sidecar
interprets) on BottleRecord; `register(..., policy=)` stores it,
`set_policy(bottle_id, policy)` updates it live, and `attribute` returns
it (the source-IP-keyed resolution).
* service.py — `launch_bottle(..., policy=)` and `set_policy`.
* control_plane.py — `POST /bottles` accepts `policy`; `PUT
/bottles/<id>/policy` live-reloads it; `POST /resolve` returns
{bottle_id, policy} for a verified (source_ip, token) — the per-request
call the multi-tenant sidecar makes; `/attribute` stays identity-only.
Scope note: this is the control-plane / state half. The data-plane half —
the egress mitmproxy addon (and git-gate) selecting allowlist / DLP /
token-injection per client IP by calling `/resolve` — is the next slice
(route agent bottles through the shared sidecar). The orchestrator stays
policy-agnostic: it stores and serves the blob verbatim.
Tests: registry policy store/update/persist; Orchestrator launch-with-policy
+ live set_policy; control-plane resolve returns policy (403 on bad token),
PUT policy updates / 404 / 400. Verified live over HTTP (launch -> resolve
-> PUT reload -> resolve reflects). Full suite green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Answers "where do we build the consolidated sidecar": nowhere, until now.
* sidecar.py — `Sidecar.ensure_built()` (default no-op) + `DockerSidecar`
now defaults its image to the real bundle (`bot-bottle-sidecars`) and
`ensure_built()` builds it from `Dockerfile.sidecars` when
`docker image inspect` shows it's missing (no-op when present or when no
dockerfile is configured, e.g. a pre-pulled image). `image_exists()`
added.
* service.py — `ensure_sidecar()` now builds then runs.
* __main__.py — `--sidecar` runs the consolidated bundle (build-if-missing).
Scope note: this builds + launches the bundle *container*; making the
running instance functional across bottles needs the per-bottle,
source-IP-keyed multi-tenant config + registration/reload, and routing
agent bottles to it — the next slices (added to PRD 0070's roadmap).
Tests: unit (docker mocked) — image_exists, ensure_built builds when
missing / no-op when present / no-op without a dockerfile / raises on build
failure; ensure_sidecar builds-then-runs; integration (gated, no heavy
build) — image_exists reflects real docker state. Full suite green (only
pre-existing /bin/sleep errors).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Review feedback: don't bury a parallel docker util in the orchestrator.
But reusing backend.docker.util (docker_mod) isn't right either — importing
it runs backend/__init__.py, which eagerly loads all three backends
(docker + firecracker + macos) plus the manifest/egress/git-gate/supervise
framework (~76 modules), so every orchestrator import would drag the whole
backend layer in.
Compromise: promote the helper to a top-level, framework-free
bot_bottle/docker_cmd.py (single stdlib import), a proper shared home the
orchestrator's docker components use now and backend.docker.util can adopt
later. Verified `import bot_bottle.orchestrator` stays lean (12 modules, no
firecracker/macos backends).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The core consolidation win: one persistent sidecar per host, shared by
every bottle, instead of a sidecar bundle per bottle. Safe to share
because the attribution invariant (source IP + identity token) lets the
sidecar map each request to the right bottle.
* orchestrator/sidecar.py — a backend-neutral `Sidecar` lifecycle
contract (mirrors LaunchBroker) + a `DockerSidecar` impl. The defining
behaviour is idempotent singleton: `ensure_running` starts the instance
if absent and is a no-op if it's already up, so N launches never spawn
N sidecars; `stop` is idempotent.
* orchestrator/dockerutil.py — a shared `run_docker` helper; DockerBroker
now uses it too (DRY with slice 3).
* service.py — the Orchestrator holds an optional `Sidecar`, exposes
`ensure_sidecar()` + `sidecar_status()`.
* control_plane.py — `GET /sidecar` reports it; __main__ gains
`--sidecar-image` and ensures the single sidecar on startup.
Tests: unit (docker mocked) — is_running, ensure idempotent (no-op when up,
starts when absent), failure raises, stop idempotent; Orchestrator sidecar
wiring/status; control-plane /sidecar; integration (gated) — ensure is a
real idempotent singleton (one container after two ensures), stop removes.
Full suite green (only pre-existing /bin/sleep errors); integration
verified locally against real docker.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The first concrete LaunchBroker, proving the orchestrator -> backend seam
on the cheapest backend (the sidecar bundle is already containers):
* orchestrator/docker_broker.py — DockerBroker runs a container on a
verified launch (`docker run --detach --name <bottle> --label ...
<image_ref>`) and removes it on teardown (`docker rm --force`,
idempotent on an already-absent container). The argv is built only from
the request's static ids/flags, so nothing free-form reaches docker;
provenance/schema verification is inherited from LaunchBroker.submit.
* __main__.py gains `--broker {stub,docker}` so the harness can drive real
containers.
Slice 3 launches a single container from image_ref (the seam); the full
agent + sidecar bundle is a later slice.
Tests: unit (docker mocked) — argv from static fields, launch/teardown call
the right commands, missing-image and docker-failure raise, teardown
idempotent on missing, forged token never touches docker; integration
(gated on a reachable daemon) — launch creates a real container, teardown
removes it. Full suite green (only pre-existing /bin/sleep errors);
integration verified locally against real docker.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Second slice of PRD 0070, still a backend-neutral dev-harness:
* orchestrator/broker.py — the launch-broker contract. A LaunchRequest is
structured (static ids/flags only — bottle id, pool slot, a
content-addressed image_ref; never a path/argv) and signed as a compact
HS256 JWT so the broker verifies PROVENANCE before acting: a compromised
co-located component can't forge a launch without the shared secret.
verify_request is fail-closed (bad sig / malformed / off-schema -> raise).
Stdlib only (no runtime deps). Ships a StubBroker that records verified
requests for the harness/tests.
* orchestrator/service.py — the Orchestrator: owns the registry and brokers
the lifecycle. launch_bottle mints the bottle + sends a signed launch,
rolling the registry entry back if the launch fails (no orphans);
teardown_bottle brokers teardown then deregisters; attribute delegates.
* control_plane.py — POST /bottles now launches, DELETE tears down (both go
through the Orchestrator + broker). dispatch/server take an Orchestrator.
* __main__.py wires an ephemeral secret + StubBroker for the harness.
Tests: broker sign/verify round-trip, tamper/wrong-secret/malformed/off-schema
rejection, StubBroker fail-closed; Orchestrator launch->registry->attribute,
teardown, rollback-on-broker-failure; control-plane updated for launch/teardown.
Full suite green (only the pre-existing /bin/sleep errors); harness does
launch -> attribute -> teardown over HTTP.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
paths is the single home now, so stop re-exporting the path helpers from
supervise: remove host_db_path from supervise's imports + __all__ (it was
re-export-only) and drop bot_bottle_root from __all__ (kept as an import,
still used by audit_dir). supervise_types was already clean. Repoint the
last readers (test_supervise imports host_db_path from paths;
test_supervise_edge calls paths.bot_bottle_root) and refresh the doc
mentions. No supervise.bot_bottle_root / supervise.host_db_path references
remain.
Behavior-preserving: full unit suite unchanged (only the pre-existing
/bin/sleep sidecar-init errors).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Create bot_bottle/paths.py as the canonical home for the app-root path
helpers (bot_bottle_root, host_db_path, HOST_DB_FILENAME) — foundational,
not supervise- or db-specific. `bot_bottle_root()` now honours a
BOT_BOTTLE_ROOT env override.
Repoint every consumer (supervise, supervise_types, db_store, queue_store,
audit_store, store_manager, bottle_state, cli/supervise, docker/cleanup,
orchestrator/registry) at paths; remove the definitions (and supervise's
duplicate host_db_path) and the now-dead `import sys`. Add paths.py to the
sidecar bundle (Dockerfile.sidecars) for the flat-import copies.
Tests: replace ~12 files' monkeypatching of supervise.bot_bottle_root (and
the flat/pkg/supervise_types triple-patch dance) with a single
`use_bottle_root()` helper that sets BOT_BOTTLE_ROOT — every module and
flat/package copy reads the same env var, so one override covers them all.
Net -97 lines. Behaviour-preserving: full unit suite unchanged (only the
pre-existing /bin/sleep sidecar-init errors remain).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
host_db_path is shared DB infrastructure, not supervise-specific, so its
canonical home is db_store (alongside DbStore). It resolves bot_bottle_root
from supervise_types lazily inside the function — no load-time cycle, and a
monkey-patch of supervise_types.bot_bottle_root still propagates.
supervise_types re-exports both names for the historical import path
(queue_store/audit_store unchanged); the orchestrator registry now imports
from db_store. Drops the now-unused `import sys` from supervise_types.
Behavior-preserving: full unit suite unchanged (only the pre-existing
/bin/sleep sidecar-init errors remain); monkeypatch propagation verified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Per review: use one shared bot-bottle.db for all runtime state including
the registry (DbStore namespaces by schema_key), so it's one queryable
file for backup/console. default_db_path() -> host_db_path(). Drop the
unilateral WAL flip — WAL on the shared DB affects supervise/audit and is
finicky over guest shares, so it's a deliberate future change; keep a
busy_timeout for lock contention.
PRD State section updated: integrity now by SOLE ownership (only the
orchestrator opens bot-bottle.db; data plane + console reach state via the
control-plane RPC, never a file handle) rather than ro/rw mount-splitting,
which one shared file can't do. Notes the transitional caveat that the
supervise sidecar currently rw-mounts bot-bottle.db.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Per review: the SQLite runtime-state DB lives on the host, not owned
inside the orchestrator unit. Durability (re-adoption must survive an
orchestrator restart) + integrity via access-scoping (control-plane rw,
data-plane ro) rather than location. Note the WAL-over-guest-share wrinkle
for the VM slices (may want a host-side DB owner reached over the RPC).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
First implementation slice of PRD 0070, the backend-neutral consolidation
core as a plain-process dev-harness (no VM packaging yet):
* orchestrator/registry.py — SQLite (WAL) runtime-state store on the
existing DbStore/TableMigrations base. Live bottle registry keyed by
source IP + per-bottle identity token, with fail-closed attribution:
a request resolves to a bottle only when its source IP AND identity
token both match exactly one active record (unknown/ambiguous IP,
empty token, or token mismatch all deny). Tokens are 256-bit urandom.
* orchestrator/control_plane.py — the HTTP control plane (the universal
transport chosen in 0070): register / deregister / list / attribute /
health. Routing is a pure dispatch() so it is socket-free testable;
Handler/ControlPlaneServer/make_server are a thin stdlib adapter.
register/deregister are the live-reload path; listing redacts tokens.
* orchestrator/__main__.py — `python -m bot_bottle.orchestrator` harness.
Launch/teardown, the launch broker, and the egress/git/supervise data
plane come in later slices. 24 unit tests (attribution matrix, persistence,
dispatch, one real-socket round-trip).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Per review: resolvable at implementation time (host tweak acceptable, as
the pool setup already needs on NixOS); the earlier sidecars-in-VMs spike
showed feasibility. No need to hunt the spike now.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Add a "Secret handling — FUTURE pattern (not v1)" subsection: vault as a
separate trust domain holding long-lived roots, deriving short-lived
scoped creds where the upstream allows (with the honest limit that a
compromised proxy can still abuse currently-authorized access). The
mechanism is a generic, user-extensible SecretProvider that generalizes
PRD 0048's DeployKeyProvisioner and drops into the manifest wherever a raw
token is accepted — discovered from ~/.bot-bottle/contrib like user
AgentProviders. Marked explicitly as not required for the initial
orchestrator; tracked as #355.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Fold the per-bottle sidecar bundle into a single persistent per-host
orchestrator: runs the sidecar functions (egress/git-gate/supervise),
coordinates with the console, and brokers agent launches. Virtualized
from the start with backend-native isolation (fc VM / apple ctr / docker
ctr), fronted by a single backend-agnostic contract; per-backend
variation lives on BottleBackend, not a parallel Orchestrator hierarchy.
Leads with the security review (secret concentration, shared fate, the
launch-broker-as-new-privileged-core, and the source-IP attribution
invariant each backend must enforce). Proposes one SQLite DB owned by the
orchestrator for runtime state (slot leases, approvals, registry) —
distinct from build-time constants (flat .env) and user config
(declarative ~/.bot-bottle). Sequences docker -> firecracker -> macos,
developing the service as a plain-process dev-harness before the VM.
Supersedes 0069's Stage-1/4 sidecar framing; depends on 0069's nix-built
fixed images. Tracking issue #351.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The pool params (size, IP base, iface prefix, nft table) were triplicated
— hardcoded in netpool.py, scripts/firecracker-netpool.sh, and
nix/firecracker-netpool.nix — plus the IP math (3x) and the nft ruleset
(2x). Nothing enforced agreement; changing the base (ce3fad9, off CGNAT)
forced a coordinated three-file edit, and a missed one would silently
provision a range the launcher doesn't expect.
Collapse to one source of truth:
* netpool.defaults.env — a plain KEY=VALUE file (bash-sourceable,
systemd EnvironmentFile-compatible, Python- and Nix-parseable) holding
the four defaults. A BOT_BOTTLE_FC_* env var still overrides any key.
* netpool.py reads it for the Python defaults (missing file = hard
error, not confusing empty defaults).
* the shell script falls back to it (no literal `:-8` / `10.243.0.0`),
and its `up` is now non-destructive/idempotent (only creates a
missing TAP), so re-running never cuts a live VM.
* the Nix module readFile-parses it for its option defaults and
delegates bring-up to the SAME shell script (dropping its duplicate
IP math, nft ruleset, and TAP loop) — passing every value as
Environment= so the store-detached script never needs the file.
Net: defaults 3x -> 1x, nft ruleset 2x -> 1x, TAP loop 2x -> 1x. The one
remaining IP-math dup (Python launch-addressing vs bash bring-up) is
justified — different runtimes. Tests now guard the invariant (Python
reads the shared file; the script/module hold no literals) instead of
pinning duplicated strings.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Split the self-hosted KVM runner setup out of this branch so main's CI
doesn't require a runner that isn't registered yet. The `coverage` job
runs on ubuntu-latest again and reports combined coverage informationally
— the hard diff-coverage gate (which needs the firecracker integration
suite, hence /dev/kvm + the provisioned pool) is deferred and moves to a
separate PR. The unit-coverage tests added alongside the gate stay.
See PRD 0069 / #348; the runner + gate re-land on the ci-runner branch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Capture the vision: drop the host Docker dependency (rootfs export +
sidecar containers) so the Firecracker backend needs only firecracker +
KVM. Staged plan (persistent per-host sidecar → nix-built fixed images →
in-VM Dockerfile builds → sidecar/builder as VMs), the docker inventory,
the secret-concentration tradeoff, and the link to rootless launch /
the paused minimal-runner work (#348).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Add a `group` option to the netpool NixOS module + BOT_BOTTLE_FC_GROUP to
the shell script: when set, the pool's TAP devices are owned by a group
instead of a single user, so any group member can open them (the kernel
lets a TAP's owning-group members attach). This lets an interactive user
and, say, a CI-runner user share one pool. `owner`/`group` are mutually
exclusive (asserted). Single-user `owner` remains the default.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The Firecracker VM/SSH orchestration (launch/boot/SSH/isolation-probe,
~230 lines) is covered by the integration suite, which needs /dev/kvm +
the provisioned pool — a container runner skips it, so those lines read
uncovered and the 90% diff gate can't pass there (it's been red since the
backend landed). Move the `coverage` job to a self-hosted `kvm` runner
with a firecracker-readiness preflight (binary + /dev/kvm + `backend
status`), so the integration test actually runs and the orchestration is
covered. Unit/lint stay on ubuntu-latest. README documents the runner
prerequisites.
Also close the last unit-coverable gaps (bottle exec/close, bottle_plan
properties, docker status/setup branches, netpool span/conflict) so the
gate clears 90% (90.3%) with the firecracker integration test running.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Cleanup orphan-enumeration + removal, the backend setup/status/teardown/
cleanup classmethod delegates, guest-IP parsing, the unprivileged nft/TAP
probes, console-tail, and require_firecracker preflight branches. Lifts
the unit-only diff-coverage on the firecracker backend substantially;
the deep VM boot/SSH orchestration remains integration-covered.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Unit tests for the generic host-setup paths across firecracker/docker/
macos-container — NixOS vs systemd (root/non-root) vs neither, missing
binary/daemon/service, persistence reporting, and the netpool shell
renderer. Lifts diff-coverage on the new setup code from ~0 to full.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The TestNetpoolOverlap `_routes` helper had an untyped fake_run + a bare
`dict`/`object` return, tripping 9 strict-mode errors (missing type args,
missing param annotations, `object` not usable as a context manager).
Type it properly: entries: list[dict[str, str]], fake_run params, and a
`AbstractContextManager[MagicMock]` return. `pyright` is clean again.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Add a "Superseded (2026-07-11)" banner and flip Status: Active →
Superseded on the 8 smolmachines-specific docs (7 PRDs + the VM-backend
research). Bodies are left intact as a historical record; the banner
points at the removal commit and the landscape doc so a reader isn't
misled into thinking the backend still exists.
Incidental mentions in other PRDs (git-gate, cred-proxy, print-parity,
etc.) are left as-is.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The control SSH logs in as root, so the agent (dropped to node via
runuser) inherited cwd=/root. That was harmless while the rootless
rootfs left /root node-owned, but the dropbear fix (chown /root →
root:root 0700) made /root unreadable to node — so the agent's cwd was
inaccessible, breaking Node's process.cwd(), Claude Code's shell-snapshot
machinery, and `/doctor` (which reported /root/.claude EACCES and failed
every Bash command).
Run the agent from its workdir (default /home/node) via `env --chdir`.
Not a `sh -c 'cd … && exec "$@"'` wrapper: ssh space-joins everything
after the host into one string for the guest shell, so the quoted script
+ $@ get re-split and mangled (it exec'd the $0 placeholder → exit 127).
`env --chdir=DIR …` is all simple words, so it survives the join.
Verified end-to-end: a real headless claude bottle now runs its Bash tool
and exits 0 (was EACCES/127 before).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Make the network pool a persistent, distro-uniform resource instead of a
non-persistent per-distro shell command. systemd is the common
denominator across Debian/Ubuntu/Fedora/RHEL/Arch/… (and NixOS), so
`backend setup` on any systemd host now installs one bot-bottle-owned
oneshot unit (bot-bottle-firecracker-netpool.service): params pinned via
Environment= (so it doesn't depend on $SUDO_USER at boot), ExecStart/Stop
delegating to the bundled bring-up script (single source of logic).
- render_systemd_unit() in netpool.py (derived from the same constants
as the shell script + nix module).
- backend setup: install + enable the unit directly when run as root,
else print a self-contained copy-paste block. NixOS keeps its
declarative module (which produces the same-named unit); non-systemd
hosts fall back to the raw imperative script.
- backend teardown: symmetric — disable + remove the unit.
- backend status: report the unit's active/inactive state so you can
tell a persistent install from an imperative one.
Net: one install path (`backend setup`), persistent, identical across
distros; per-distro variance shrinks to installing nft + iproute2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
First real end-to-end launch (there was no firecracker integration test)
surfaced three bugs in the control/provision path, all now verified fixed
by tests/integration/test_firecracker_launch:
1. Guest SSH rejected the correct key. The rootless rootfs build
(`docker export | tar` as a non-root user) can't preserve uid 0, so
every path — including /root — is owned by the build uid (node in
guest). dropbear refuses root's authorized_keys when /root isn't
root-owned, so auth fell back to password → denied. bb-init now
`chown -R 0:0 /root`.
2. The SSH client (newer OpenSSH) didn't reliably present the -i key
against the operator's ~/.ssh config; add `IdentitiesOnly=yes` to
ssh_base_argv so only the per-bottle key is offered.
3. cp_in double-wrapped the remote command as `sh -c <remote>`, but ssh
space-joins everything after the host into one string for the guest
shell, collapsing `sh -c mkdir -p X && …` to `mkdir` with no operand
("missing operand"). Pass the command as a single arg and let the
guest login shell run it (stdin still carries the tar).
Also stop discarding dropbear's stderr (`-E` now reaches the host-side
console.log) so future guest-auth issues are debuggable.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
`status()` hard-failed when it couldn't confirm the nft table, but listing
nftables usually needs root — so an unprivileged `backend status` reported
"not ready" even with the pool fully up, making it useless as a launch
gate (and skipping the firecracker integration test on a set-up host).
Base readiness on what the launch preflight actually hard-requires: the
TAP pool present (unprivileged, authoritative) + no range overlap. Report
the nft table state (present / unverified / not-confirmable-unprivileged)
but don't let it flip readiness — the post-boot isolation probe is the
authoritative isolation check, same as the preflight's deferral.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The module used `networking.nftables.tables.*` (which forces
`networking.nftables.enable = true`, flipping the host firewall backend)
and `systemd.network.enable` (handing interfaces to systemd-networkd) —
disruptive on an iptables + Docker daily driver.
Rewrite it to a single systemd oneshot that brings the pool up: idempotent
`ip tuntap` for the TAPs + `nft -f` for the independent `inet bot_bottle_fc`
table (its own hooks at priority -10), with ExecStop teardown. Same as the
imperative script, but declarative. It touches neither the firewall backend
nor networkd, so it coexists with iptables/Docker/ufw/firewalld. Verified
through the NixOS module system (service present, firewall untouched).
Drop the now-redundant `render_nixos_module()` paste generator (the module
is a real importable file) and point `backend setup` at importing it (flake
output or the file path), noting it's non-invasive.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
`backend setup --backend=firecracker` only covered the network pool. Lead
with the other prerequisites: the firecracker binary (with a release
install pointer when it's not on PATH, plus a NixOS note), a /dev/kvm
check, and a reminder about the cached guest kernel + static dropbear and
mke2fs. The network pool is now step 2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Add an abstract teardown() classmethod to BottleBackend — the inverse of
setup(), surfaced as `./cli.py backend teardown [--backend=NAME]`
(uninstall). Symmetric with setup: it prints the privileged commands /
declarative config change to remove the host prerequisites.
- firecracker: NixOS-aware — disable the flake module (or drop the
import) and rebuild, or `firecracker-netpool.sh down` imperatively.
- docker / macos-container: nothing to undo (no privileged host state);
print a short note.
Not called by the launch path or the test suite. Extends test_cli_backend
for the new dispatch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Sandbox-escape couldn't run: its git-gate fixture had no host_key, so the
preflight ssh-keyscanned the deliberately-unreachable upstream and die()d
in setUpClass. Preset a throwaway host_key so the keyscan is skipped (the
key is never used — the push is rejected by gitleaks first).
That unblocked a second, pre-existing issue: the planted secrets weren't
caught by gitleaks (the AWS example key is allowlisted; the others hit
entropy/keyword gates in the keyword-free URL the attack embeds). Reshape
the fixtures — three structural, high-entropy shapes gitleaks matches
without a keyword (github / slack / gitlab) for the git-push attack, and
a separate alphanumeric secret for the DNS attack (a gitleaks-matchable
token carries separators that aren't valid DNS labels, so the two uses
can't share one secret). All five sandbox-escape tests now pass.
Add test_firecracker_launch: a launch smoke (exec + proxy env) gated on
`FirecrackerBottleBackend.status() == 0` (the `backend status` result),
skipping with setup instructions when the TAP pool / nft table aren't
provisioned.
The git-gate-only-matches-gitleaks-patterns asymmetry the reshape exposed
is tracked separately (#346).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Add abstract `setup()`/`status()` classmethods to BottleBackend (same
per-host, no-instance shape as is_available) so host provisioning is
part of the backend contract, not a per-backend command. Replace the
`./cli.py firecracker {setup,status}` command with a generic
`./cli.py backend {setup,status} [--backend=NAME]` that resolves a
backend (flag / $BOT_BOTTLE_BACKEND / host default) and dispatches —
swapping backends is just a different --backend.
Implementations:
- firecracker: moved out of cli/ into backend/firecracker/setup.py
(network pool module/script + range-overlap check), unchanged output.
- docker: new backend/docker/setup.py — reports docker on PATH, daemon
reachability, and gVisor runsc; setup notes no privileged pool is
needed. Minimal placeholder; richer version tracked in #345.
- macos-container: new setup.py — container CLI + system-service checks.
Also retarget the launch-preflight / isolation-probe pointers to the new
command. New test_cli_backend covers dispatch + docker setup/status.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
The default TAP-pool base was 100.64.0.0/10 (RFC-6598 CGNAT) — chosen to
dodge RFC-1918, but that's exactly the range Tailscale assigns node
addresses from, so on a Tailscale host it's the worst pick. Move the
default to 10.243.0.0/16, an obscure RFC-1918 block that steers clear of
docker/libvirt/k8s/LAN and Tailscale.
No default is collision-proof, so add netpool.overlapping_routes(): it
parses `ip -json route show table all` and flags any route intersecting
the pool range (excluding our own bbfc* TAPs and the default route). The
launch preflight warns on overlap; `backend status` reports it.
Distribute the NixOS host setup as a flake module instead of a
copy-pasted blob: nix/firecracker-netpool.nix computes the taps / nft
table from typed options (poolSize, ipBase, ifacePrefix, owner) with a
/31-alignment assertion, and flake.nix exposes it as
nixosModules.firecracker-netpool. Defaults mirror the backend constants;
writeEnvFile emits the matching BOT_BOTTLE_FC_* so the host pool and the
launcher can't drift.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Delete the smolmachines backend (the whole bot_bottle/backend/smolmachines
package and its tests). It had fatal Linux issues (TSI networking under
sustained use, exec-channel contention, no SIGWINCH) and is superseded by
the Firecracker backend (issue #342).
Backend selection now:
- default is macos-container on macOS, firecracker on KVM-capable Linux
hosts, and docker as the last resort (was smolmachines).
- firecracker is selected on a KVM host even when the `firecracker`
binary isn't installed, so start routes through its preflight and
prints an install pointer (same UX as require_container), instead of
silently falling back. Split is_host_capable() (Linux + KVM) out of
is_available() (adds the binary check) to drive this.
Retarget the cross-backend tests (parity, print-parity, prepare,
workspace, freezer, selection) from smolmachines to firecracker rather
than dropping the coverage. Remove docker.util.image_id/save, which only
smolmachines used. Update README/AGENTS/example bottles and stale
comments; historical docs/prds are left as a point-in-time record.
BREAKING: BOT_BOTTLE_BACKEND=smolmachines now errors as unknown.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WBMWTEtQdJ4W5UrWuLHCck
Adds a Firecracker-based backend for Linux, providing mature KVM-based
microVM isolation to replace smolmachines/libkrun (issue #342, closes
the dead-end tracked in #332).
Architecture:
- Guest control over SSH (dropbear injected into the rootfs) on a
point-to-point TAP link. `ssh -t` forwards SIGWINCH natively, so no
resize bridge is needed.
- Networking: a one-time, root-provisioned pool of user-owned TAP
devices (no shared bridge → no docker0/virbr0/cni0 collisions) plus a
dedicated `table inet bot_bottle_fc` nftables table (independent of
Docker/ufw/firewalld rules). `./cli.py firecracker setup` prints the
host-appropriate config (NixOS module or sudo script).
- Rootfs: `docker export` → ext4 via `mke2fs -d` (rootless, no mount),
cached by image digest; per-bottle SSH pubkey + IP passed via the
kernel cmdline.
- Sidecar: reuses the Docker bundle, published on the slot's host TAP IP.
- Fail-closed isolation: TAP pool verified at preflight; the egress
boundary is proven empirically post-boot (before the agent runs) by a
canary probe — the VM must fail to reach the host directly, or launch
is refused.
Linux hosts with Firecracker + KVM now default to this backend;
macOS stays on macos-container.
Not yet validated end-to-end on live hardware (requires the one-time
network pool). Unit tests + pyright pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01G8p32HJgPoS1hLPWubbftM
- Broaden scope from "Claude Code in Docker" to "AI coding agents in isolated sandboxes"
- Add SuperHQ (superhq.ai, v0.4.4) as a new adjacent-competitor entry: macOS-only
microVM desktop app, overlaps on isolation/credential-proxy/multi-provider but has
no manifest layer and no audit logging
- Note SuperHQ's user-voiced audit gap (Brian Cheong, Dunialabs.io) and that
bot-bottle already covers it
- Update differentiation list to reflect three backends (Docker, Apple container,
smolmachines) and three built-in providers (Claude Code, Codex, Pi) plus plugin system
- Add in-flight directions for forge-native dispatch (#317) and paid web control plane
(#327) with honest framing: lifecycle concept is not novel vs. cloud services (Devin,
Copilot Workspace); differentiation is self-hosted + manifest-driven + stronger isolation
CI runs as root, so chmod 0o000 doesn't prevent reads. Both
test_skips_unreadable_file (TestFindRepoBottleFile) and
test_returns_false_when_file_unreadable_on_write now mock
Path.read_text to raise OSError directly.
DX improvement: find_repo_bottle_file() is called before the save
prompt so the user sees exactly which file will be updated:
Save host_key for 'myrepo' to /home/.../bottles/dev.md? [y/N]
If no bottle file is found the prompt is skipped and a clear message
explains that the key is session-only.
Also:
- Use find_repo_bottle_file() inside find_and_update_bottle_file()
instead of duplicating the scan logic
- Rename all module-level helpers to public names (no underscore prefix)
since they are imported and tested externally
The regex-based _insert_host_key_in_frontmatter had a bug: child_indent
was overwritten on every line with indent > repo_indent (including
grandchildren), so host_key landed inside the key: block instead of at
the repo level.
Replace with a clean parse → mutate → re-serialize approach:
- Add serialize_yaml_subset() to yaml_subset.py (block-style, 2-space
indent, round-trips through parse_yaml_subset)
- Replace _insert_host_key_in_frontmatter + _update_frontmatter_in_file
with _add_host_key_to_frontmatter that parses the frontmatter dict,
sets fm["git-gate"]["repos"][name]["host_key"], and re-serializes
- _find_and_update_bottle_file simplified to read file → delegate to
_add_host_key_to_frontmatter → write if changed
didericis
Quality Badge Bot
didericis-claude